Rome's Gothic Wars: From the Third Century to Alaric (Key Conflicts of Classical Antiquity)

by Michael Kulikowski

Gothic Migration in the Archaeological Evidence

  But does the identification allow us to do more than that? For instance, does the identification of the Sântana-de-Mureş/Černjachov culture with fourth-century Goths allow us to find Goths elsewhere? Many archaeologists and historians would answer yes. The argument has been made most explicitly by Volker Bierbrauer: the Sântana-de-Mureş/Černjachov archaeological culture is Gothic; some of its characteristics – particular brooch and ceramic types, a tendency not to place weapons in graves – are similar to those of the Wielbark culture, which was centred on the Vistula river and lasted from the first to the fourth century A.D.; the Wielbark culture must therefore also be Gothic. Also, because the Sântana-de-Mureş/Černjachov culture is Gothic, and because some artefacts associated with it appear inside the frontiers of the Roman empire, these artefacts must represent the movement of Goths from the Danube to Italy, and thence to Gaul and Spain. Bierbrauer’s simplistic ethnic ascription model is extreme, but only because it is articulated so clearly.[47] Unfortunately, many other archaeologists and historians working in the field accept its core assumptions without acknowledging the fact. Even Peter Heather, the most subtle modern interpreter of Gothic history, has written about ‘working backwards’ from the Sântana-de-Mureş/Černjachov culture to earlier stages of ‘Gothic’ archaeology.[49] Two separate considerations, one practical, one theoretical, make this approach untenable.

  For one thing, the Sântana-de-Mureş/Černjachov culture is extremely diverse. As we shall see in the next chapter, the artefacts, construction techniques, and burial practices found within the Sântana-de-Mureş/Černjachov zone have parallels with earlier cultural traditions within the zone itself, with Roman provincial culture, with the Wielbark and Przeworsk cultures to the north and west, and with the steppe cultures of the east. The Wielbark elements in the Sântana-de-Mureş/Černjachov culture are no more numerous than other elements, so there is no archaeological reason to privilege them over others. Even if Wielbark artefacts were dominant in the Sântana-de-Mureş/Černjachov zone, they would not necessarily signify the same thing in both places: artefacts that are emblematic of one thing in one place may change meaning radically if transposed to another. More importantly still, the closeness of the artefactual connections between the two cultures is not as great as is usually asserted. Indeed, their chief point of intersection is not particular artefacts, but the fact that weapon burials are absent from the Wielbark and rare in the Sântana-de-Mureş/Černjachov zones. In purely logical terms, a negative characteristic is less convincing proof of similarity than a positive one, and the fact that weapon burials are commonest where archaeological investigation has been most intensive suggests that our evidentiary base is anything but representative. Given this, why should the Wielbark–Sântana-de-Mureş/Černjachov connection seem so self-evident to so many scholars? One answer is an old methodology that seeks to explain changes in material culture by reference to migration. The other is Jordanes.

  Migration v. Diffusion Theories

  The methodological problem is of long standing. In the early years of archaeology’s development as a scientific discipline, it was normal to understand cultural changes as the result of one tribe or people conquering or displacing another and replacing the previous material culture with a new one of their own. This interpretative paradigm goes back in part to the nationalist scholarship of the Volk at which we have already looked, in part to the preoccupation of our ancient historical sources with invasion, migration and conquest, and in part to Kossinna’s ascription of fixed and defined material cultures to ethnopolitical groupings. In the 1970s and 1980s, some archaeological theorists reacted radically against such migration theories. Working from the simple and obvious observation that the material culture of a place can change radically without the population of that place changing much at all, these archaeologists sought to explain change in archaeological cultures by reference to the diffusion of materials and ideas rather than migration. Diffusionist theory became and remains the norm, particularly amongst British archaeologists.

  On the other hand, diffusionist theory, like any theory, can be pushed to unrealistic extremes. It is, after all, a simple fact that people move and have always done so, sometimes over long distances – a fine example from our period is the Sarmatian Iazyges, who moved en masse from the vicinity of the Dnieper and Dniester rivers, where Strabo places them at the beginning of the first century A.D., to the Alföld between the Danube and Tisza, where Pliny places them in the 70s A.D., having come at the request of the Quadic king Vannius for aid against the Hermunduri. When people move, they often bring large parts of their native culture with them, however transformed it may be when transplanted into a new environment: one need only look at any large immigrant neighbourhood in the U.S. or Britain to see the truth of this fact. What is more, the conquest of one region by people from another can profoundly alter the culture of a conquered region, with or without massive population shifts: the expansion of the Roman empire is history’s best illustration of this. Each of these points contradicts the more extreme statements of radical diffusionist theory, but it is unfortunate that this kind of overstatement has given comfort to those who would rather think solely in terms of migration and conquest. The truth of the matter, as so often, lies in the middle ground. Massive cultural changes can take place without much movement of population; by the same token, large-scale movements of population have obviously taken place in the past, which means that some massive cultural changes should indeed be explicable in terms of migration. Neither migration nor diffusion will suit every case, neither can be denied in every case, and we should always have a reason for asserting one explanation over the other in any given instance.

  The deep attachment to migration theories in the case of the Goths – and the reading of connections between Wielbark and Sântana-de-Mureş/Černjachov cultures in terms of Gothic migration – can be explained without any deep engagement with archaeological theory. The reading of both Wielbark and Sântana-de-Mureş/Černjachov cultures is what we might call ‘text-hindered’ and Jordanes is the culprit.[50] His migration story takes the Goths from Scandinavia to the Baltic and then to the Black Sea. Archaeologists have therefore been called upon not to read the material evidence on its own terms, but rather to prove or disprove the authenticity of Jordanes’ text. In 1970, Rolf Hachmann disproved the Scandinavian connection on archaeological grounds, thereby making necessary new theories of ethnogenesis such as we have looked at earlier.[51] But the question has remained the same for the Baltic–Black Sea sequence: can one prove or disprove Jordanes? For an archaeologist of the Goths like Michel Kazanski, this is not even a question: the text of Jordanes tells us the Goths were at the Baltic, then in the Ukraine; therefore the material culture of both regions must be Gothic and we should study it as such.[52] That is precisely what we mean when we say the topic is text-hindered: consciously or not, the archaeological question is always structured by Jordanes, hence an insistence on drawing out the material similarities between the Wielbark and the Sântana-de-Mureş/Černjachov cultures.

  If we did not have Jordanes, that connection would not seem self-evident. Taken on purely archaeological grounds, without reference to our one piece of textual evidence, there is no reason to interpret the Wielbark and the Sântana-de-Mureş/Černjachov cultures as close cousins. The Sântana-de-Mureş/Černjachov culture represents an intermingling of many different earlier material cultures, some native to its zone, others not. One might argue, as most do, that the Sântana-de-Mureş/Černjachov culture came into being because of a migration out of the Wielbark regions, but one might equally argue that it was an indigenous development of local Pontic, Carpic and Dacian cultures or of the migration of steppe nomads from the east meeting Przeworsk-culture warriors from the west. In purely archaeological terms, each of these interpretations is equally possible, for as we have seen, Wielbark cultural elements are no more numerous in the Sântana-de-Mureş/Černjachov culture than are
the many other cultural traditions that make it up. It is only the text of Jordanes that leads scholars to privilege the Wielbark connection. Indeed, if Jordanes did not exist and we were dealing with truly prehistoric cultures, it is highly unlikely that anyone would draw the same connection.

  How the Sântana-de-Mureş/Černjachov Culture Became Gothic

  What, then, are we to make of all this? How are we to interpret the origins of the Sântana-de-Mureş/Černjachov culture and the Gothic hegemony with which it coincides chronologically? Is there such a thing as Gothic history before the third century? The answer, at least in my view, is that there is no Gothic history before the third century. The Goths are a product of the Roman frontier, just like the Franks and the Alamanni who appear at the same time. That is clearly demonstrated by contemporary literary evidence, and indeed all the evidence of the fourth and fifth centuries – everything except the sixth-century Jordanes. In the third century, the Roman empire was assaulted from the regions north of the Danube and the Black Sea by large numbers of different barbarian groups, among whom Goths appear for the first time. Not long thereafter, the Goths are clearly the most powerful group in the region, while most of the other barbarian groups with whom they appear in the third century either disappear from the record or are clearly subordinated to them. The most plausible explanation of this evidence is to see one group among the many different barbarians north of the Black Sea establishing its hegemony over the scattered and hitherto disparate population of the region, which was thereafter regularly identified as Gothic by Graeco-Roman observers.

  The archaeological evidence of the Sântana-de-Mureş/Černjachov culture makes sense in these terms as well. The rise to prominence of a few strong leaders created a stable political zone in which a single material culture came into being, synthesized from a variety of disparate traditions. None was more important than the others – as the material evidence clearly shows – and there is no need to look for ‘original’ Goths coming from elsewhere to impose their leadership and their identity on others. There were, of course, immigrants into the region where the Sântana-de-Mureş/Černjachov culture arose, from elsewhere in northern and central Europe and from the steppe lands to the east as well. But none of them need themselves have been Goths, because there is no good evidence that Goths existed before the third century.

  What Made a Goth a Goth and How Can We Tell?

  That leads us back to the sense of collective identity, the problem of telling the difference that we looked at earlier. How was it that these different people knew that they were Goths rather than something else, or did they? How did Greeks and Romans know it? What marked them off as such? In most cases, context alone would have supplied the clues. There may well have been items of emblematic clothing that established insider and outsider status. But that does not mean we can construct a Gothic costume on the basis of grave finds, because in most circumstances, these items were displayed to other Goths and communicated information about status within the community, not about relations to those outside it. Language probably made a difference, and when Gothic was codified as a written religious language in the fourth century, the use of the Gothic bible will surely have identified its user as a Goth as well as a Christian. But languages can be acquired and many of the philologically Germanic languages spoken in central Europe were mutually intelligible. We have no sources to tell us that specifically Gothic idioms or accents could be used to tell a Goth from a Gepid on the Danube frontier – perhaps they could not. What was it, then, that created a sense of community among the Goths of the later third and the fourth centuries? How was it that they knew what their Greek and Roman observers claim to know – that all these people were Goths?

  It is possible that precisely the same Roman elite discourse that is accessible to us nowadays helped cultivate a sense of barbarian collective identity along different stretches of the frontier. Just as contact with the Roman empire shaped, and sometimes created, new social and political hierarchies beyond the frontier, so too Roman ideologies and perceptions may have helped single out elements in the culture of the barbarians that came to define those barbarians’ own sense of community. In other words, Roman elite discourses about what a Goth was helped to define how people came to identify themselves as Goths, to codify the signs that conveyed Gothicness. This possibility is not as strange as it might seem at first glance, as post-colonial studies of more recent periods have shown. Modern imperialism has had profound effects in shaping the identity of indigenous and subject peoples – it has been shown, for instance, that the codification of a Sikh cultural, as opposed to religious, identity was largely the result of the British need to have a readily identifiable collective group who could be employed in the colonial army.[53] That a parallel process took place along the frontiers of the Roman empire is actually quite plausible: the diverse small groups whom the Romans called Franks or Goths because they lived in a particular place and were recruited into particular units of the Roman army eventually became Franks and Goths because that was how they were described when they had political dealings with the Roman empire, when, for instance, they were recruited into Roman military units or were defeated by an emperor and described in an imperial victory title. As leaders whom Romans identified as Goths grew in strength and their followers grew in numbers, those followers became more like each other, spurred to it by the military intercourse with the empire next door. If one wants to, it is possible to call this transformation ‘ethnogenesis’ – new Gothic polities clearly came into being at the end of the third and the start of the fourth century. But it needs no appeal to Gothic aristocrats or royal lines, nor to ethnic traditions or processes, to explain what happened, and whether these new polities were very aware of being a gens or an ethnos is not something that the evidence can tell us.

  The barbaricum had always been a vast and changing place when viewed from the Graeco-Roman perspective. Probably its changeability was fully evident to those who lived in it as well. People moved about in that changing world, and alliances shifted repeatedly, sometimes at a great distance from the Roman frontier where neither Greeks and Romans nor we can have any inkling of precise circumstances. Sometimes we see tiny faded traces of changing patterns of alliance, changing patterns of trade and interaction, often no more than a shift in the routes along which Roman coins and luxury goods were dispersed. In the third century, in the region northwest of the Black Sea, the warrior stratum of a heterogeneous population came together to take advantage of imperial civil war and to reap a harvest of as much loot as speed and violence would permit. By the end of the third century, a few of these warriors were powerful enough to coordinate political control over stretches of territory north of the Danube and Black Sea. Sometimes they fought the empire, sometimes they fought each other, sometimes they served the empire, sometimes they came together and acted for their common interest. At their centre were leaders who were seen to be Goths by the Romans and who perhaps saw themselves as Goths as well. Certainly, in time, after being told repeatedly that they were in fact Goths and leaders of Gothic gentes with whom the empire would fight and make treaties, there was no question in anyone’s mind that they were indeed Goths. Likewise the Sântana-de-Mureş/Černjachov culture must surely be the result of a political stability of long enough standing for stable cultural relations to develop. That stability is attested by the growing political sophistication of the Gothic leaders whom we meet in the course of the fourth century and who form the subject of the next chapter.

  Chapter 4 Imperial Politics and the Rise of Gothic Power

  Our attempt at explaining Gothic origins has taken us a very long way from our narrative, indeed a long way from the ancient world, and into a discussion of modern intellectual history. The detour has been important. It looked at the way modern accounts of Gothic migration, whether they claim to be supported by historical, archaeological or linguistic evidence, are all in one way or another echoes of Jordanes’ sixth-century Getica. Consciously or not,
modern narratives of Gothic migration are rooted in the very old quest for Germanic origins, a quest to give northern Europe a past independent of Roman history. Unfortunately, as we have seen, contemporary evidence supports neither migration stories nor any narrative derived from Jordanes. On the contrary, it suggests that – like the Franks and the Alamanni further west along the frontier – the Goths were a product of the Roman frontier itself. That conclusion not only makes sense of the evidence of the late third century, it also fits in well with the much better understood evidence of the fourth century.

  In the first three decades of the fourth century, as we shall see in this and the next chapter, the Goths became the indisputable masters of the lower Danube, from the eastern edge of the Carpathians to the fringes of the Caucasian steppelands. Language itself began to acknowledge these facts. Thus, by the 320s, the lower Danube was known as the ripa Gothica, the Gothic bank. Soon thereafter, we find the Greek word Gothia designating the tract of land beyond the Danube, a word that was imported into the Gothic language as Gutthiuda, the Goths’ word for their own lands. This tremendous extension of Gothic power was not inevitable. Instead, the Goths were encouraged to become so powerful because it was useful to the political schemes of successive Roman emperors for them to do so. In other words, just as the Goths themselves were created by the political pressures of life in a Roman frontier zone, so Roman emperors made the fourth-century Goths what they were. The revolutionary reign of Diocletian marks the turning point.

  Map 3. The Roman Empire of Diocletian.

  Map 4. Asia Minor, the Balkans and the Black Sea region, showing Roman cities and Sântana-de-Mureş/Černjachov sites mentioned in the text.